Help us help you: an unreadable plate is a problem that started three days ago

There is a type of plate that any microbiologist recognizes instantly. You pull it from the incubator and you know there is nothing to be done with it. Confluent growth from edge to edge; more than three hundred colonies piled on top of each other with no separation; fungi that have colonized the plate so thoroughly they have formed their own Pangea, which delimits the extent of each colony by a fine darker line.

That plate cannot be counted. Not by eye, not with a magnifying glass, not with the best image analysis system that exists. And the reason lies in the inoculation and incubation process.

It may seem trivial, but we tend to think of counting as the critical moment, the one that requires attention and judgment. But counting can only be as good as the plate in front of you. A poorly planned inoculation is equivalent to a bad result that is impossible to rescue, a plate that was never designed to be counted.

In this article, we will discuss just that: how to inoculate so that the resulting plate can be read. It is not a minor or obvious topic. It is, probably, the part of the process where the quality of the final result is most decided.

Not all inoculation techniques are for counting

Before discussing how to inoculate well for counting, it is worth clarifying something that in practice generates more confusion than it should: not all inoculation techniques are designed for counting. Each one has a different purpose, and using one for something it was not designed for is a source of problems in itself.

Streak plating (or streak plate for isolation) has a very specific objective: to isolate individual colonies from a mixture, separating one strain from another or purifying a culture. The zigzag pattern, with the loop flamed between sectors, progressively reduces the density of cells until in the final sector well-separated colonies grow that can be picked and identified. It is an isolation technique, not a counting technique. The colonies it produces are not quantifiable because the amount of inoculum deposited is neither known nor reproducible. A plate inoculated by streaking is not a plate to be counted — it is a plate from which selections are made.

Spread plating (surface inoculation) is indeed a counting technique. A known volume of diluted sample, typically 0.1 mL, is deposited and distributed homogeneously over the surface of the agar with a Drigalski spreader or equivalent. Colonies grow on the surface, where they are visible, accessible, and easy to distinguish. It is the inoculation method of choice when you want to count aerobic microorganisms and when you want colonies to be large and well-defined.

Pour plating (depth inoculation) is the other major counting technique. A known volume of sample is mixed with melted and tempered agar before it solidifies. Colonies grow both on the surface and within the agar mass. It allows counting somewhat larger sample volumes and is the basis of ISO 4833-1, but it has a downside: internal colonies are smaller and sometimes more difficult to visualize than surface colonies.

Lawn plating seeks the opposite of counting: confluent and uniform growth across the entire plate. It is the basis of disk diffusion antibiograms and phage titration by plaque assays. Here the objective is not to count colonies, but to have a continuous layer on which to measure inhibition halos or conduct plaque assays. Counting a lawn makes no sense — it is not made for that.

This separation of purposes is well established in methodological literature: streaking is described for isolating individual colonies, while pour plating and spread plating are reserved for enumerating viable colonies, and soft agar overlays for titrating phages (Sanders, 2012, Journal of Visualized Experiments). These are not interchangeable variants of the same technique, but tools for different problems.

The distinction matters because it determines which plate makes sense to take to a count and which does not. A plate from streak isolation, a lawn plate, or a contaminated plate are not failed counting plates: they are plates that never had the intention to be counted. Quantitative counting lives in spread plating and pour plating on solid medium. Everything else solves other problems.

The error of rushing

Of all inoculation errors, there is one that is committed more than any other and that almost always has the same cause: impatience.

After spreading the inoculum on the surface, you must wait for it to be completely absorbed before inverting the plate for incubation. If you invert it while the inoculum is still wet, the liquid shifts by gravity, drags cells to one side, and ruins the uniform distribution you just achieved. The result is a plate with colonies piled up in one zone and empty in another. Impossible to count in a representative way.

The wait is about two minutes. Just two minutes. But it is surprisingly difficult to convince someone in a hurry that it is worth it, because two minutes standing in front of the plate feels like wasted time when fifty samples are waiting. They are not wasted. It is the difference between a readable plate and one that will have to be redone, and redoing costs far more than two minutes.

It is the error of rushing par excellence, and the solution is not technical. It is understanding why the wait exists. When you understand that wet inoculum is free liquid that will move as soon as you tilt the plate, and with it the distribution of the sample.

Getting the dilution right before you gamble

The most common problem behind an uncountable plate, more than three hundred colonies, confluent growth, is almost always the same: the wrong dilution was inoculated because the starting concentration of the sample was unknown.

When working with a sample of unknown concentration, inoculating a dilution directly and waiting to see what happens is a gamble. If you get it right, perfect. If not, you have wasted plate, medium, incubation time, and sample to end up with something that cannot be counted, and you have to start over.

There are ways to reduce that risk. The most elegant one I know is the Miles and Misra method, which allows you to visualize the gradient of several serial dilutions on a single plate before committing to a full inoculation. Originally described in 1938 as a method to estimate the bactericidal power of blood, it has survived nearly a century precisely because its counts fit with remarkable fidelity to a Poisson distribution, which makes it statistically sound despite its simplicity (Miles, Misra, and Irwin, 1938, Journal of Hygiene). The plate is divided into sectors, a small drop (typically 20 µL) of each serial dilution is deposited in its sector, and allowed to grow. The next day, one of the dilutions will have fallen into the optimal density — separated and countable colonies — and that tells you exactly which dilution to inoculate on a full plate to get a result within range.

It has a cost: it requires a prior night of incubation, and the sample must withstand refrigeration until then without its microbial load changing significantly. It is not always viable. But when it is, it saves an enormous amount of wasted plates and failed counts. It helped me tremendously when I was starting out, because it turns the choice of dilution into an informed decision rather than a gamble.

When Miles and Misra is not practical, the reasonable alternative is to always inoculate two or three consecutive dilutions in parallel. It multiplies the number of plates, yes, but it guarantees that at least one will fall within the countable range, and it avoids the complete repetition of the analysis three days later. Between wasting two extra plates today and repeating the entire assay on Friday, the math is clear.

Inoculate with the person who will count in mind

A good inoculation for counting does not just seek to grow something. It seeks to make what grows easy to read. And "easy to read" has concrete requirements.

The first is correct density: enough colonies for the result to be statistically representative, but not so many that they touch and fuse. The countable range of the standard (15 to 300 colonies according to ISO, 25 to 250 according to other guidelines) makes sense. It is the goal of inoculation. When you inoculate, you are trying to land within that range. Everything you do with the dilution is directed toward that.

The second is uniform distribution. A plate with colonies spread homogeneously can be counted at a glance, can be divided into sectors if needed, and allows reliable extrapolation. A plate with colonies huddled on one side is not representative even if the total number is in range. Homogeneous spreading of the inoculum and waiting for it to dry are what produce that uniformity.

The third is separation between colonies, and here is a point that is underestimated. Colonies that touch are a problem not only for the human eye, but especially for any image-based reading. When two colonies fuse, neither an analyst nor a camera can know for certain whether it was one large colony or two small ones stuck together. With fungi this becomes critical: excessive fungal growth, where colonies have extended until they touch, leaves no space to distinguish individual units. The plate becomes unreadable not because there is too much, but because there are no longer boundaries between things.

That is why an inoculation designed for counting is, fundamentally, an inoculation that respects space. One that aims for a density where each colony has room to grow isolated from its neighbors. This facilitates counting, whatever the method, because they all depend on the same thing: being able to trace the border of each colony without ambiguity.

Reading conditions help, but they do not rescue. Good indirect lighting, appropriate contrast background, and a clean plate underneath improve reading of a well-inoculated plate. They do not fix a poorly inoculated one. Light does not separate colonies that grew stuck together.

Counting begins in the inoculation

Going back to the plate from the beginning, the confluent one, the one with fungi touching each other, the one with three hundred colonies piled on top of each other.

The analyst who pulls it from the incubator did not make any error. The error was already made. It was in the dilution that was chosen without knowing the starting concentration, in the inoculum that was inverted before it dried, or in the lack of a test inoculation that would have warned that this sample had much higher load than expected.

This is what I want to convey: counting is not an isolated act at the end of the process. It is the final stage of a chain that begins with inoculation, and the quality of the result is decided long before anyone sits down to count. A well-inoculated plate almost counts itself. A poorly inoculated plate cannot be counted in any way.

The next time a plate comes out uncountable, the useful question is not "how do I count it." It is "what happened in the inoculation." Almost always, the answer is there.

References

• Miles, A.A., Misra, S.S., and Irwin, J.O. (1938). The estimation of the bactericidal power of the blood. Journal of Hygiene, 38(6), 732–749. — Original description of the drop method for viable count on surface. https://doi.org/10.1017/S002217240001158X

• Sanders, E.R. (2012). Aseptic laboratory techniques: plating methods. Journal of Visualized Experiments (JoVE). — Different purposes of streaking (isolation), pour plating and spread plating (enumeration), and soft agar overlay (phage titration). https://doi.org/10.3791/3064

• ISO 4833-1:2013 and ISO 4833-2:2013. Microbiology of the food chain — Horizontal method for the enumeration of microorganisms. International Organization for Standardization. — Depth plating and surface plating techniques, countable range. https://www.iso.org/standard/53728.html and https://www.iso.org/standard/59509.html